Enteroscopy-assisted endoscopic retrograde cholangiopancreatography (EA-ERCP) is an established technique for managing benign hepaticojejunostomy (HJ) stenosis. Compared to alternative interventions, such as percutaneous or endosonography-guided approaches, EA-ERCP is less invasive and offers a favorable safety profile. However, technical success rates of EA-ERCP may be lower in anatomically complex cases, such as those involving Roux-en-Y (RY) anastomoses, as compared to simpler BII gastroenteric anastomoses used in pancreatoduodenectomy (PD). Additionally, it is currently unclear whether balloon dilatation alone, or combined with stenting, represents the most effective endoscopic treatment strategy. The primary study aim was to evaluate the efficacy and safety of EA-ERCP for suspected benign HJ stenosis in patients with RY versus PD. The secondary aim was to assess the outcomes of balloon dilatation with or without stent placement.  

This single-center retrospective study included patients who underwent EA-ERCP for suspected benign HJ stenosis between 2018 and 2025. Eligible patients had previously undergone PD or HJ with RY, while excluding those with post-gastric bypass anatomy. Procedures were performed using a short- or long-type double-balloon enteroscope. HJ stenosis was treated with balloon dilatation alone, or when the stricture failed to resolve adequately, balloon dilatation was combined with stent therapy. A metal covered Nagi stent was placed if technically possible, if not, multiple plastic stents were inserted. Recurrences of HJ stenosis were generally managed with repeat balloon dilatation combined with stent therapy.                                                                                                                                                                           The primary outcome was technical success. Secondary outcomes included clinical success (symptom resolution 14 days post-EA-ERCP), adverse events (AGREE classification), recurrent HJ stenosis and the need for reintervention. 

A total of 192 patients underwent 320 EA-ERCPs. Benign HJ stenosis was confirmed in 102 of 192 (53.1%) index EA-ERCPs. Overall technical success rate was 279/320 (87.2%); 194/220 (88.2%) in PD group and 85/100 (85.0%) in RY group [p=0.16]. Technical failures in PD versus RY patients included inability to reach the HJ (12; 5,5% vs 13; 13,0%), to canulate the bile ducts (13; 5,9% vs 2; 2,0%) and to complete the intended therapy (1; 0,5% vs none). Overall clinical success was 83,9% (PD 116/135 (85,9%) versus RY 35/45 (77,8%), p=0.35). Overall adverse event (AE) rate was 16/320 (5.0%) (PD 11/220; 5.0% versus RY 5/100; 5.0%), without severe (grade IV/V) AEs. At index EA-ERCP, 79 patients were treated with balloon dilatation. Recurrence of HJ stenosis requiring reintervention was observed in 37/79 (46.8%) patients at a median of 8 months [IQR: 1.5-29], which occurred in 25/37 (67.6%) cases in the PD group and 12/37 (32.4%) patients in the RY group. Median total follow-up was 17.0 months [IQR: 8.0-35.3]. HJ patency at 1, 2 and 3 years was 67.9%, 58.7% and 40.4%, respectively. Stent therapy was initiated in 33 patients, either by multiple plastic stents (n=24) or a Nagi stent (n=9). Stent dysfunction requiring reintervention occurred in 5/24 (20.8%) patients treated with plastic stents and in 4/9(44%) patients with a Nagi stent. Recurrence of HJ stenosis for which reintervention was needed, was observed in 8/33 (26%) patients following stenting therapy, after a median follow up of 8 months [IQR 1-14]. Recurrence of HJ stenosis after stent therapy in PD group was 4/8 (50%) and 4/8 (50%) in RY group.

EA-ERCP for suspected HJ stenosis shows a similar high technical and clinical success rate in patients with PD and RY anatomy with a favorable safety profile. These findings reinforce the role of EA-ERCP as first line approach in benign HJ stenosis in both types of anatomical reconstructions. However, future studies should focus on optimizing the therapeutic strategy to reduce recurrences of HJ stenosis.